Grantee Research Project Results
Final Report: Early Career Award: Framework for Quantifying Microbial Risk and Sustainability of Potable Reuse Systems in the United States
EPA Grant Number: R835823Title: Early Career Award: Framework for Quantifying Microbial Risk and Sustainability of Potable Reuse Systems in the United States
Investigators: Gerrity, Daniel
Institution: University of Nevada - Las Vegas
EPA Project Officer: Packard, Benjamin H
Project Period: August 1, 2015 through July 31, 2018
Project Amount: $329,650
RFA: Human and Ecological Health Impacts Associated with Water Reuse and Conservation Practices (2014) RFA Text | Recipients Lists
Research Category: Water , Human Health
Objective:
This research addressed three critical issues related to the broad implementation of potable reuse in the United States: (1) developing quantitative microbial risk assessments (QMRAs) associated with various indirect potable reuse (IPR) and direct potable reuse (DPR) treatment trains; (2) developing a framework for comparing the sustainability of IPR versus DPR using Las Vegas as a case study; and (3) evaluating ozone-biofiltration for its ability to mitigate the potential public health implications of disinfection byproducts (DBPs) and other contaminants of emerging concern (CECs) in advanced treated wastewaters. The overall objective of the research was to assess whether ‘planned’ IPR and DPR systems, which typically incorporate advanced water purification with ozone-biofiltration and/or reverse osmosis (RO), offer safe and sustainable alternatives to ‘conventional’ drinking water. Many ‘conventional’ drinking water systems are influenced by upstream wastewater discharges but do not specifically address the potential increase in contaminant loadings with dedicated advanced treatment processes—a practice commonly described as de facto reuse.
Summary/Accomplishments (Outputs/Outcomes):
Task 1. This task was primarily focused on static and dynamic QMRA models of representative bacterial (Salmonella), viral [adenovirus (AdV) and norovirus (NoV)], and protozoan (Cryptosporidium) pathogens. The research was divided into three subtasks involving the development and analysis of (1) a static QMRA for Cryptosporidium, (2) a static QMRA for the entire suite of target pathogens, and (3) a dynamic QMRA for NoV. All three models were developed using a system dynamics approach. (Model 1) Due to the robust nature of planned IPR and the resiliency offered by the environmental buffer, the risk of cryptosporidiosis was not particularly sensitive to upsets in the engineered treatment train. Instead, major variables affecting risk included Cryptosporidium concentrations in the upstream source water, dilution ratios (i.e., recycled water contributions), and storage time and temperature in the environmental buffer. For the baseline condition (270 days of storage time in the environmental buffer at a temperature of 20°C), the risks associated with de facto reuse and planned IPR were identical, with the upstream surface water controlling the risk for both systems. In fact, annual risk decreased with increasing recycled water contribution for both de facto reuse and planned IPR. The advanced treated water in the planned IPR system was actually quite superior to the de facto reuse system, but the risk increased significantly after blending in the reservoir. For the ‘critical condition’ (i.e., 105 days of storage time in the environmental buffer at a temperature of 10°C), the risk for de facto reuse sometimes exceeded that of conventional drinking water, while the planned IPR system was sufficiently robust to still achieve the 10-4 annual risk benchmark in all instances. instances. The risks associated with DPR were always lower than the 10-4 annual risk benchmark, even during failure scenarios. (Model 2) The annual risks of infection for the DPR treatment trains and for IPR with full advanced treatment (FAT) and groundwater replenishment were typically less than the benchmark risk of 10-4 for each individual pathogen and also for the combined pathogen risk. The only exception was for AdV in the DPR system employing FAT and direct distribution (maximum annual risk = 7.3×10-4), which was due to AdV’s resistance to UV disinfection. However, the UV process was conservatively modeled with a dose of 80 mJ/cm2, so more typical UV advanced oxidation doses (i.e., >100 mJ/cm2) are expected to achieve the 10-4 benchmark. All of the treatment trains employing either surface water discharge or blending exhibited identical results for annual risk for all pathogens, with values exceeding the 10-4 annual risk benchmark. This indicates that the risks for these treatment trains were dominated by pathogen concentrations in the upstream surface water. Sensitivity analyses for outbreak conditions, storage time/temperature, and treatment process failures were also used to identify ‘critical’ conditions. (Model 3) Based on a comparison of the static and dynamic NoV models, the dynamic model resulted in notably higher risks. In fact, estimated risks for the de facto reuse system increased by two orders of magnitude, and risks in the FAT-based DPR system increased by eight orders of magnitude. The major differences in model structure included (1) varying primary transmission based on disease incidence within the community; (2) secondary transmission, which was expected to increase risk; and (3) the distributed delays associated with the epidemiological states, which were expected to decrease risk. Subsequent model runs indicated that dynamic primary transmission had a negligible impact on cumulative incidence (CI) and that secondary transmission was primarily driving risk in the potable reuse systems. Overall, NoV risk was lower than the 10-4 risk benchmark (even during a foodborne outbreak), except for the de facto reuse systems and when considering endemic disease within the community and/or secondary transmission.
Task 2: The uncertain hydrologic conditions of the Colorado River system potentially offer justification for future implementation of DPR in vulnerable communities. Nevada recently revised its regulations to permit IPR via groundwater replenishment, but DPR has not yet been regulated in Nevada. Nevertheless, with recent momentum in surrounding states, DPR could eventually become a reality if a need was demonstrated. As such, the goal of this task was to present a framework for evaluating the sustainability of IPR versus DPR (with alternative treatment trains) and to determine whether DPR is a viable option for Las Vegas considering its site-specific factors. The results of a system dynamics model highlighted some of the limitations of Las Vegas’ water supply framework that could be mitigated through DPR implementation, but it also identified limitations that must be overcome to make DPR a viable option. Task 1 indicated that treatment trains employing reverse osmosis (RO) or ozone-biofiltration are generally ‘equivalent’ in the context of public health protection, and the Task 2 model indicated that both could achieve net increases in water supply relative to the status quo RFC approach. The RO-based train could even achieve improvements in water quality, specifically related to total dissolved solids (TDS) and trace organic compound (TOrC) concentrations in finished drinking water. But despite the fact that DPR with direct distribution could reduce energy consumption, greenhouse gas (GHG) emissions, and salt and/or phosphorus loadings to Lake Mead, both treatment trains would likely be cost-prohibitive relative to the status quo approach. The net present worth of DPR ranged from $1.5 billion to $4.7 billion (in 2015 dollars) depending on treatment train, diversion percentage, and use of the product water (i.e., raw water vs. treated water augmentation), while the net present worth of the status quo RFC approach was ~$1.0 billion. The ozone-biofiltration treatment train would also be hindered by unacceptably high TDS concentrations (up to 1,500 mg/L)—and possibly TOrC concentrations—in the finished drinking water. That being said, potable reuse is highly site-specific, with varying regulatory frameworks, levels of public acceptance, and practical considerations impacting its implementation in different cities. Even though DPR may not be a viable alternative for southern Nevada, this research presented a framework for evaluating DPR in other locations. In locations with low source water TDS concentrations, the cheaper ozone-biofiltration treatment train may produce an acceptable drinking water quality, even without the use of RO. Moreover, in locations with even greater elevation changes, the energy cost differential coupled with a cost efficient treatment train may yield an economically viable alternative.
Task 3. This task employed a pilot-scale ozone-biofiltration system treating full-scale membrane bioreactor (MBR) filtrate from a municipal water reclamation facility. This task focused on quantifying total organic carbon (TOC) removal and DBP formation potential using the uniform formation conditions (UFC) approach for free chlorine [targeting trihalomethanes (THMs) and haloacetic acids (HAAs)] and chloramines [targeting N-nitrosodimethylamine (NDMA)]. Direct NDMA formation and attenuation were also assessed. The performance of the system was evaluated based on operational conditions, including ozone dose and empty bed contact time (EBCT), and media type [anthracite vs. biological activated carbon (BAC)]. The experiments yielded empirical kinetics models for predicting TOC removal as a function of O3/TOC ratio, EBCT, and media type. The free chlorine UFC experiments also yielded an empirical correlation between THM and HAA formation potential and effluent TOC concentration in the ozone-biofiltration effluent. Using this relationship, the research identified a target effluent TOC concentration of 2 mg/L for compliance with DBP regulations in the United States. The direct NDMA formation/attenuation experiments identified a threshold EBCT of 10 minutes for NDMA removal with either anthracite or BAC. Interestingly, NDMA removal appeared to be more dependent on the established microbial community in the biofiltration columns than short-term changes in operational conditions, specifically dissolved oxygen (DO) and biodegradable dissolved organic carbon (BDOC) levels. Data from 16S rRNA gene sequencing confirmed differences in microbial community structure as a function of pre-ozonation and media type, but additional microbiological testing is needed to identify the specific bacteria responsible for NDMA biodegradation and the conditions promoting their colonization of biofiltration systems (e.g., pre-ozonated feed water). The chloramines UFC experiments demonstrated that biofiltration alone was ineffective in reducing NDMA formation potential (3% reduction) but that ozonation was highly effective (96% reduction). Post-ozone biofiltration reduced NDMA even further, but it is unclear whether the biofiltration step removed additional precursors or eliminated the NDMA that had formed during ozonation. Finally, TOrC analyses generally supported previously published literature and indicated that perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA) (combined concentration of ~25 ng/L) may not be a concern for this particular system (U.S. EPA Health Advisory Level = 70 ng/L).
Task 4. The key accomplishment was the production of a short film describing the water cycle in Las Vegas, specifically in the context of potable reuse. The effort was integrated into a documentary filmmaking course at UNLV, and the students in the course were responsible for all aspects of production, including research, script writing, hiring of actors/actresses, filming, narrating, and video and sound editing. The team incorporated a virtual tour of the Las Vegas water cycle and street interviews with the public, including interviews on the Las Vegas Strip, on the campus of UNLV, and at Hoover Dam. The film students also met with water and wastewater professionals from local industry, including representatives from the Southern Nevada Water Authority (SNWA) and Clark County Water Reclamation District (CCWRD). Two students involved in the development of the short film received an Emmy nomination for Student Craft-Directing and an Emmy award for Student Craft-Writing from the National Academy of Television Arts & Sciences Pacific Southwest Chapter.
Journal Articles on this Report : 3 Displayed | Download in RIS Format
Other project views: | All 23 publications | 3 publications in selected types | All 3 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Amoueyan E, Ahmad S, Eisenberg JNS, Pecson B, Gerrity D. Quantifying pathogen risks associated with potable reuse: a risk assessment case study for Cryptosporidium. Water Research 2017;119:252–266. |
R835823 (2016) R835823 (2017) R835823 (Final) |
Exit Exit Exit |
|
Arnold M, Batista J, Dickenson E, Gerrity D. Use of ozone-biofiltration for bulk organic removal and disinfection byproduct mitigation in potable reuse applications. Chemosphere 2018;202:228-237. |
R835823 (Final) |
Exit Exit Exit |
|
Gerrity D, Arnold M, Dickenson E, Moser D, Sackett JD, Wert EC. Microbial community characterization of ozone-biofiltration systems in drinking water and potable reuse applications. Water Research 2018;135:207-219. |
R835823 (Final) |
Exit Exit Exit |
Supplemental Keywords:
drinking water, decision making, engineeringRelevant Websites:
The following websites contain information related to this EPA-funded research. The first is an episode of UNLV Research Files highlighting the project, and the second is a short film produced in collaboration with the UNLV Department of Film.
UNLV Research Files Gerrity Exit
Progress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.